X-Message-Number: 11664
From: Thomas Donaldson <>
Subject: More re computer emulation with responses to some posters
Date: Sun, 2 May 1999 22:31:52 +1000 (EST)

Hi everyone!

Well, let's see. 

Several messages for this Cryonet (sent out for 2 May 1999) appear to
believe that emulation of reality is easy. One major point I made in my
posting was that emulation is NOT easy, especially since we do not know
reality completely and therefore cannot emulate it. Unexpected events 
happen all about us. That is Reality. And to me it seems quite unlikely
that we will EVER completely understand the Universe.

(A full emulation of the Universe, of course, might take a VERY large
computer --- too large to exist in the Universe. But I think of that
as a subsidiary issue).

I also asked a question, seriously: if you believe that a character
living entirely in a computer and receiving only input from the computer
is aware, then do you also believe that characters in computer games are
aware? If not why not; if you do, please explain. Just what is the 
difference between such characters and any more elaborate program
structure in a computer?

Some have also suggested that we might already be such an emulation. I
find this possibility not so much wrong as meaningless: sure, so we are
emulations. And just what does that really tell us? The idea has no
experimental content at all, just like a belief in God. 

Another participant points out that we can have dreams which feel like
reality. Fine. Since we have those dreams with our own biological brains,
and they feel like reality (but afterwards we see that they were not)
this has no obvious relation to the possibility of emulating a person
totally inside a computer. IF you assume that we can achieve the same
result with a computer, then it proves what you want --- but the argument
here is precisely about whether any such assumption is correct.

And finally I note that Damien Broderick has suggested that we consider
not simply computers with no means to sense the outside world, but a 
computer which CAN sense and respond to the outside world. That's fine.
There are many more difficulties to actually doing that than some 
computer people may realize, but I myself would agree that if we add
those features then we are not longer talking about computer emulations
anyway. Welcome to the club. I will point out that there are physical
limits to just how fast a sequential computer can work (limits for
quantum computation or parallel computation will clearly be different,
though they also exist). Even to be able to sense the world in real
time as well as human beings do, a device must work much faster than any
sequential computer. (You might be able to make a sequential beetle,
though; and if not that, a sequential worm. They have few enough neurons
that you might be able to get something which works like them in real
time).

Fundamentally, I have no problems with the idea that we might create a
device (note that I do not say "computer") able to be aware and respond
to the world as well as we do (including making mistakes as we do). Such
a device will have to be highly parallel (just as we are) and therefore
not well-modelled by a Turing machine.

		Best wishes and long long life to all,

			Thomas Donaldson

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=11664